I've been thinking about this whole "Religion is a modern, western, concept and therefore it is anachronistic and illicit to think in terms of 'religion' in pre-modern, pre-western, contexts." It occurs to me that this is grounded in a fundamental misconception of what it means for something to be socially constructed, and this due to an antecedent commitment to what Lonergan calls "idealism."
Five difficulties. at least, are overlooked in this discussion. First, they elide the distinction between "constructed" and "not real." Social and linguistic constructions are too often treated as if they are ipso facto false when in fact the exact opposite is the case. Constructions are very real, with very real consequences. Just ask the people dying in Iraq because ISIS has constructed them as polytheists. Second, they elide the distinction between "similar" and "sameness." To call Islam a religion and Buddhism a religion is not to say that they are the same but rather that in certain ways they are similar, perhaps most notably a concern with what we might call "ultimate reality" (i.e. what is the ultimate origin of all that we experience; and pointing out that Buddhism's answer is to deny the reality of ultimate reality would hardly mean that Buddhism is unconcerned with ultimate reality but merely to state its own distinctive approach to the matter). Third, they entail a markedly artificial understanding of discourse, one that ignores how people actually use language. The supposition is that if in reference to a given historical context we cannot speak about religion in the sense that we use it in reference to contemporary phenomenon then we cannot speak about religion at all. Under-considered is the reality that when in academic discourse I use "early Christian religion" I recognize, and assume reasonably that competent readers will recognize, that this is something quite distinct from "modern Christian religion," that in fact the adjectives "early" and "modern" do important linguistic work, signalling the precise semantic meaning to be given to both "Christian" and "religion." Fourth, it fails to consider that just because people are unaware that something is happening in their midst does not mean that it was not happening in their midst. Meyer talks about this with reference to development: Christian religion was always marked by development, even if it took until the mid-19th century for Christians to develop the conceptual tools to recognize that development. This is to say, showing that the ancients had no concept that translates directly to what we mean by "religion" does not mean that they had no phenomena that could be grouped under that term as the broad umbrella. Fifth, there is lurking in the background an interesting western and modernist linguistic hegemony: if the pre-modern or a non-western people does not have religion as we understand the term then they did not or do not have religion at all.
Ultimately much of this discussion flounders on the rocks of pragmatism. We have to have some word to distinguish analytically between what an ancient Jewish women did daily in the marketplace and what she did when she went on pilgrimage to Jerusalem. If we excise "religion" and by extension the adjective "religious" then we will have to invent terms that do the work that they did. Rather than reinvent the wheel why not simply retain those terms and remember the hardly new lesson that their religion is not the same as our religion? I.e. that to group two distinct entities under the categorical heading "religion" denotes that they are similar not that they are the same.